1
0
Fork 0
alistair23-linux/arch/arm64/include/asm/assembler.h

390 lines
8.9 KiB
C
Raw Normal View History

/*
* Based on arch/arm/include/asm/assembler.h, arch/arm/mm/proc-macros.S
*
* Copyright (C) 1996-2000 Russell King
* Copyright (C) 2012 ARM Ltd.
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#ifndef __ASSEMBLY__
#error "Only include this from assembly code"
#endif
#ifndef __ASM_ASSEMBLER_H
#define __ASM_ASSEMBLER_H
#include <asm/asm-offsets.h>
#include <asm/cpufeature.h>
#include <asm/page.h>
#include <asm/pgtable-hwdef.h>
#include <asm/ptrace.h>
arm64: debug: avoid accessing mdscr_el1 on fault paths where possible Since mdscr_el1 is part of the debug register group, it is highly likely to be trapped by a hypervisor to prevent virtual machines from debugging (buggering?) each other. Unfortunately, this absolutely destroys our performance, since we access the register on many of our low-level fault handling paths to keep track of the various debug state machines. This patch removes our dependency on mdscr_el1 in the case that debugging is not being used. More specifically we: - Use TIF_SINGLESTEP to indicate that a task is stepping at EL0 and avoid disabling step in the MDSCR when we don't need to. MDSCR_EL1.SS handling is moved to kernel_entry, when trapping from userspace. - Ensure debug exceptions are re-enabled on *all* exception entry paths, even the debug exception handling path (where we re-enable exceptions after invoking the handler). Since we can now rely on MDSCR_EL1.SS being cleared by the entry code, exception handlers can usually enable debug immediately before enabling interrupts. - Remove all debug exception unmasking from ret_to_user and el1_preempt, since we will never get here with debug exceptions masked. This results in a slight change to kernel debug behaviour, where we now step into interrupt handlers and data aborts from EL1 when debugging the kernel, which is actually a useful thing to do. A side-effect of this is that it *does* potentially prevent stepping off {break,watch}points when there is a high-frequency interrupt source (e.g. a timer), so a debugger would need to use either breakpoints or manually disable interrupts to get around this issue. With this patch applied, guest performance is restored under KVM when debug register accesses are trapped (and we get a measurable performance increase on the host on Cortex-A57 too). Cc: Ian Campbell <ian.campbell@citrix.com> Tested-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
2014-04-29 12:04:06 -06:00
#include <asm/thread_info.h>
/*
* Enable and disable interrupts.
*/
.macro disable_irq
msr daifset, #2
.endm
.macro enable_irq
msr daifclr, #2
.endm
/*
* Enable and disable debug exceptions.
*/
.macro disable_dbg
msr daifset, #8
.endm
.macro enable_dbg
msr daifclr, #8
.endm
arm64: debug: avoid accessing mdscr_el1 on fault paths where possible Since mdscr_el1 is part of the debug register group, it is highly likely to be trapped by a hypervisor to prevent virtual machines from debugging (buggering?) each other. Unfortunately, this absolutely destroys our performance, since we access the register on many of our low-level fault handling paths to keep track of the various debug state machines. This patch removes our dependency on mdscr_el1 in the case that debugging is not being used. More specifically we: - Use TIF_SINGLESTEP to indicate that a task is stepping at EL0 and avoid disabling step in the MDSCR when we don't need to. MDSCR_EL1.SS handling is moved to kernel_entry, when trapping from userspace. - Ensure debug exceptions are re-enabled on *all* exception entry paths, even the debug exception handling path (where we re-enable exceptions after invoking the handler). Since we can now rely on MDSCR_EL1.SS being cleared by the entry code, exception handlers can usually enable debug immediately before enabling interrupts. - Remove all debug exception unmasking from ret_to_user and el1_preempt, since we will never get here with debug exceptions masked. This results in a slight change to kernel debug behaviour, where we now step into interrupt handlers and data aborts from EL1 when debugging the kernel, which is actually a useful thing to do. A side-effect of this is that it *does* potentially prevent stepping off {break,watch}points when there is a high-frequency interrupt source (e.g. a timer), so a debugger would need to use either breakpoints or manually disable interrupts to get around this issue. With this patch applied, guest performance is restored under KVM when debug register accesses are trapped (and we get a measurable performance increase on the host on Cortex-A57 too). Cc: Ian Campbell <ian.campbell@citrix.com> Tested-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
2014-04-29 12:04:06 -06:00
.macro disable_step_tsk, flgs, tmp
tbz \flgs, #TIF_SINGLESTEP, 9990f
mrs \tmp, mdscr_el1
bic \tmp, \tmp, #1
msr mdscr_el1, \tmp
arm64: debug: avoid accessing mdscr_el1 on fault paths where possible Since mdscr_el1 is part of the debug register group, it is highly likely to be trapped by a hypervisor to prevent virtual machines from debugging (buggering?) each other. Unfortunately, this absolutely destroys our performance, since we access the register on many of our low-level fault handling paths to keep track of the various debug state machines. This patch removes our dependency on mdscr_el1 in the case that debugging is not being used. More specifically we: - Use TIF_SINGLESTEP to indicate that a task is stepping at EL0 and avoid disabling step in the MDSCR when we don't need to. MDSCR_EL1.SS handling is moved to kernel_entry, when trapping from userspace. - Ensure debug exceptions are re-enabled on *all* exception entry paths, even the debug exception handling path (where we re-enable exceptions after invoking the handler). Since we can now rely on MDSCR_EL1.SS being cleared by the entry code, exception handlers can usually enable debug immediately before enabling interrupts. - Remove all debug exception unmasking from ret_to_user and el1_preempt, since we will never get here with debug exceptions masked. This results in a slight change to kernel debug behaviour, where we now step into interrupt handlers and data aborts from EL1 when debugging the kernel, which is actually a useful thing to do. A side-effect of this is that it *does* potentially prevent stepping off {break,watch}points when there is a high-frequency interrupt source (e.g. a timer), so a debugger would need to use either breakpoints or manually disable interrupts to get around this issue. With this patch applied, guest performance is restored under KVM when debug register accesses are trapped (and we get a measurable performance increase on the host on Cortex-A57 too). Cc: Ian Campbell <ian.campbell@citrix.com> Tested-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
2014-04-29 12:04:06 -06:00
isb // Synchronise with enable_dbg
9990:
.endm
arm64: debug: avoid accessing mdscr_el1 on fault paths where possible Since mdscr_el1 is part of the debug register group, it is highly likely to be trapped by a hypervisor to prevent virtual machines from debugging (buggering?) each other. Unfortunately, this absolutely destroys our performance, since we access the register on many of our low-level fault handling paths to keep track of the various debug state machines. This patch removes our dependency on mdscr_el1 in the case that debugging is not being used. More specifically we: - Use TIF_SINGLESTEP to indicate that a task is stepping at EL0 and avoid disabling step in the MDSCR when we don't need to. MDSCR_EL1.SS handling is moved to kernel_entry, when trapping from userspace. - Ensure debug exceptions are re-enabled on *all* exception entry paths, even the debug exception handling path (where we re-enable exceptions after invoking the handler). Since we can now rely on MDSCR_EL1.SS being cleared by the entry code, exception handlers can usually enable debug immediately before enabling interrupts. - Remove all debug exception unmasking from ret_to_user and el1_preempt, since we will never get here with debug exceptions masked. This results in a slight change to kernel debug behaviour, where we now step into interrupt handlers and data aborts from EL1 when debugging the kernel, which is actually a useful thing to do. A side-effect of this is that it *does* potentially prevent stepping off {break,watch}points when there is a high-frequency interrupt source (e.g. a timer), so a debugger would need to use either breakpoints or manually disable interrupts to get around this issue. With this patch applied, guest performance is restored under KVM when debug register accesses are trapped (and we get a measurable performance increase on the host on Cortex-A57 too). Cc: Ian Campbell <ian.campbell@citrix.com> Tested-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
2014-04-29 12:04:06 -06:00
.macro enable_step_tsk, flgs, tmp
tbz \flgs, #TIF_SINGLESTEP, 9990f
disable_dbg
mrs \tmp, mdscr_el1
orr \tmp, \tmp, #1
msr mdscr_el1, \tmp
arm64: debug: avoid accessing mdscr_el1 on fault paths where possible Since mdscr_el1 is part of the debug register group, it is highly likely to be trapped by a hypervisor to prevent virtual machines from debugging (buggering?) each other. Unfortunately, this absolutely destroys our performance, since we access the register on many of our low-level fault handling paths to keep track of the various debug state machines. This patch removes our dependency on mdscr_el1 in the case that debugging is not being used. More specifically we: - Use TIF_SINGLESTEP to indicate that a task is stepping at EL0 and avoid disabling step in the MDSCR when we don't need to. MDSCR_EL1.SS handling is moved to kernel_entry, when trapping from userspace. - Ensure debug exceptions are re-enabled on *all* exception entry paths, even the debug exception handling path (where we re-enable exceptions after invoking the handler). Since we can now rely on MDSCR_EL1.SS being cleared by the entry code, exception handlers can usually enable debug immediately before enabling interrupts. - Remove all debug exception unmasking from ret_to_user and el1_preempt, since we will never get here with debug exceptions masked. This results in a slight change to kernel debug behaviour, where we now step into interrupt handlers and data aborts from EL1 when debugging the kernel, which is actually a useful thing to do. A side-effect of this is that it *does* potentially prevent stepping off {break,watch}points when there is a high-frequency interrupt source (e.g. a timer), so a debugger would need to use either breakpoints or manually disable interrupts to get around this issue. With this patch applied, guest performance is restored under KVM when debug register accesses are trapped (and we get a measurable performance increase on the host on Cortex-A57 too). Cc: Ian Campbell <ian.campbell@citrix.com> Tested-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
2014-04-29 12:04:06 -06:00
9990:
.endm
arm64: debug: avoid accessing mdscr_el1 on fault paths where possible Since mdscr_el1 is part of the debug register group, it is highly likely to be trapped by a hypervisor to prevent virtual machines from debugging (buggering?) each other. Unfortunately, this absolutely destroys our performance, since we access the register on many of our low-level fault handling paths to keep track of the various debug state machines. This patch removes our dependency on mdscr_el1 in the case that debugging is not being used. More specifically we: - Use TIF_SINGLESTEP to indicate that a task is stepping at EL0 and avoid disabling step in the MDSCR when we don't need to. MDSCR_EL1.SS handling is moved to kernel_entry, when trapping from userspace. - Ensure debug exceptions are re-enabled on *all* exception entry paths, even the debug exception handling path (where we re-enable exceptions after invoking the handler). Since we can now rely on MDSCR_EL1.SS being cleared by the entry code, exception handlers can usually enable debug immediately before enabling interrupts. - Remove all debug exception unmasking from ret_to_user and el1_preempt, since we will never get here with debug exceptions masked. This results in a slight change to kernel debug behaviour, where we now step into interrupt handlers and data aborts from EL1 when debugging the kernel, which is actually a useful thing to do. A side-effect of this is that it *does* potentially prevent stepping off {break,watch}points when there is a high-frequency interrupt source (e.g. a timer), so a debugger would need to use either breakpoints or manually disable interrupts to get around this issue. With this patch applied, guest performance is restored under KVM when debug register accesses are trapped (and we get a measurable performance increase on the host on Cortex-A57 too). Cc: Ian Campbell <ian.campbell@citrix.com> Tested-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
2014-04-29 12:04:06 -06:00
/*
* Enable both debug exceptions and interrupts. This is likely to be
* faster than two daifclr operations, since writes to this register
* are self-synchronising.
*/
.macro enable_dbg_and_irq
msr daifclr, #(8 | 2)
.endm
/*
* SMP data memory barrier
*/
.macro smp_dmb, opt
dmb \opt
.endm
/*
* Emit an entry into the exception table
*/
.macro _asm_extable, from, to
.pushsection __ex_table, "a"
.align 3
.long (\from - .), (\to - .)
.popsection
.endm
#define USER(l, x...) \
9999: x; \
_asm_extable 9999b, l
/*
* Register aliases.
*/
lr .req x30 // link register
/*
* Vector entry
*/
.macro ventry label
.align 7
b \label
.endm
/*
* Select code when configured for BE.
*/
#ifdef CONFIG_CPU_BIG_ENDIAN
#define CPU_BE(code...) code
#else
#define CPU_BE(code...)
#endif
/*
* Select code when configured for LE.
*/
#ifdef CONFIG_CPU_BIG_ENDIAN
#define CPU_LE(code...)
#else
#define CPU_LE(code...) code
#endif
/*
* Define a macro that constructs a 64-bit value by concatenating two
* 32-bit registers. Note that on big endian systems the order of the
* registers is swapped.
*/
#ifndef CONFIG_CPU_BIG_ENDIAN
.macro regs_to_64, rd, lbits, hbits
#else
.macro regs_to_64, rd, hbits, lbits
#endif
orr \rd, \lbits, \hbits, lsl #32
.endm
/*
* Pseudo-ops for PC-relative adr/ldr/str <reg>, <symbol> where
* <symbol> is within the range +/- 4 GB of the PC.
*/
/*
* @dst: destination register (64 bit wide)
* @sym: name of the symbol
* @tmp: optional scratch register to be used if <dst> == sp, which
* is not allowed in an adrp instruction
*/
.macro adr_l, dst, sym, tmp=
.ifb \tmp
adrp \dst, \sym
add \dst, \dst, :lo12:\sym
.else
adrp \tmp, \sym
add \dst, \tmp, :lo12:\sym
.endif
.endm
/*
* @dst: destination register (32 or 64 bit wide)
* @sym: name of the symbol
* @tmp: optional 64-bit scratch register to be used if <dst> is a
* 32-bit wide register, in which case it cannot be used to hold
* the address
*/
.macro ldr_l, dst, sym, tmp=
.ifb \tmp
adrp \dst, \sym
ldr \dst, [\dst, :lo12:\sym]
.else
adrp \tmp, \sym
ldr \dst, [\tmp, :lo12:\sym]
.endif
.endm
/*
* @src: source register (32 or 64 bit wide)
* @sym: name of the symbol
* @tmp: mandatory 64-bit scratch register to calculate the address
* while <src> needs to be preserved.
*/
.macro str_l, src, sym, tmp
adrp \tmp, \sym
str \src, [\tmp, :lo12:\sym]
.endm
/*
* @sym: The name of the per-cpu variable
* @reg: Result of per_cpu(sym, smp_processor_id())
* @tmp: scratch register
*/
.macro this_cpu_ptr, sym, reg, tmp
adr_l \reg, \sym
mrs \tmp, tpidr_el1
add \reg, \reg, \tmp
.endm
/*
* vma_vm_mm - get mm pointer from vma pointer (vma->vm_mm)
*/
.macro vma_vm_mm, rd, rn
ldr \rd, [\rn, #VMA_VM_MM]
.endm
/*
* mmid - get context id from mm pointer (mm->context.id)
*/
.macro mmid, rd, rn
ldr \rd, [\rn, #MM_CONTEXT_ID]
.endm
/*
* read_ctr - read CTR_EL0. If the system has mismatched
* cache line sizes, provide the system wide safe value
* from arm64_ftr_reg_ctrel0.sys_val
*/
.macro read_ctr, reg
alternative_if_not ARM64_MISMATCHED_CACHE_LINE_SIZE
mrs \reg, ctr_el0 // read CTR
nop
alternative_else
ldr_l \reg, arm64_ftr_reg_ctrel0 + ARM64_FTR_SYSVAL
alternative_endif
.endm
/*
* raw_dcache_line_size - get the minimum D-cache line size on this CPU
* from the CTR register.
*/
.macro raw_dcache_line_size, reg, tmp
mrs \tmp, ctr_el0 // read CTR
ubfm \tmp, \tmp, #16, #19 // cache line size encoding
mov \reg, #4 // bytes per word
lsl \reg, \reg, \tmp // actual cache line size
.endm
/*
* dcache_line_size - get the safe D-cache line size across all CPUs
*/
.macro dcache_line_size, reg, tmp
read_ctr \tmp
ubfm \tmp, \tmp, #16, #19 // cache line size encoding
mov \reg, #4 // bytes per word
lsl \reg, \reg, \tmp // actual cache line size
.endm
/*
* raw_icache_line_size - get the minimum I-cache line size on this CPU
* from the CTR register.
*/
.macro raw_icache_line_size, reg, tmp
mrs \tmp, ctr_el0 // read CTR
and \tmp, \tmp, #0xf // cache line size encoding
mov \reg, #4 // bytes per word
lsl \reg, \reg, \tmp // actual cache line size
.endm
/*
* icache_line_size - get the safe I-cache line size across all CPUs
*/
.macro icache_line_size, reg, tmp
read_ctr \tmp
and \tmp, \tmp, #0xf // cache line size encoding
mov \reg, #4 // bytes per word
lsl \reg, \reg, \tmp // actual cache line size
.endm
/*
* tcr_set_idmap_t0sz - update TCR.T0SZ so that we can load the ID map
*/
.macro tcr_set_idmap_t0sz, valreg, tmpreg
#ifndef CONFIG_ARM64_VA_BITS_48
ldr_l \tmpreg, idmap_t0sz
bfi \valreg, \tmpreg, #TCR_T0SZ_OFFSET, #TCR_TxSZ_WIDTH
#endif
.endm
/*
* Macro to perform a data cache maintenance for the interval
* [kaddr, kaddr + size)
*
* op: operation passed to dc instruction
* domain: domain used in dsb instruciton
* kaddr: starting virtual address of the region
* size: size of the region
* Corrupts: kaddr, size, tmp1, tmp2
*/
.macro dcache_by_line_op op, domain, kaddr, size, tmp1, tmp2
dcache_line_size \tmp1, \tmp2
add \size, \kaddr, \size
sub \tmp2, \tmp1, #1
bic \kaddr, \kaddr, \tmp2
9998:
.if (\op == cvau || \op == cvac)
alternative_if_not ARM64_WORKAROUND_CLEAN_CACHE
dc \op, \kaddr
alternative_else
dc civac, \kaddr
alternative_endif
.else
dc \op, \kaddr
.endif
add \kaddr, \kaddr, \tmp1
cmp \kaddr, \size
b.lo 9998b
dsb \domain
.endm
/*
* reset_pmuserenr_el0 - reset PMUSERENR_EL0 if PMUv3 present
*/
.macro reset_pmuserenr_el0, tmpreg
mrs \tmpreg, id_aa64dfr0_el1 // Check ID_AA64DFR0_EL1 PMUVer
sbfx \tmpreg, \tmpreg, #8, #4
cmp \tmpreg, #1 // Skip if no PMU present
b.lt 9000f
msr pmuserenr_el0, xzr // Disable PMU access from EL0
9000:
.endm
/*
* copy_page - copy src to dest using temp registers t1-t8
*/
.macro copy_page dest:req src:req t1:req t2:req t3:req t4:req t5:req t6:req t7:req t8:req
9998: ldp \t1, \t2, [\src]
ldp \t3, \t4, [\src, #16]
ldp \t5, \t6, [\src, #32]
ldp \t7, \t8, [\src, #48]
add \src, \src, #64
stnp \t1, \t2, [\dest]
stnp \t3, \t4, [\dest, #16]
stnp \t5, \t6, [\dest, #32]
stnp \t7, \t8, [\dest, #48]
add \dest, \dest, #64
tst \src, #(PAGE_SIZE - 1)
b.ne 9998b
.endm
/*
* Annotate a function as position independent, i.e., safe to be called before
* the kernel virtual mapping is activated.
*/
#define ENDPIPROC(x) \
.globl __pi_##x; \
.type __pi_##x, %function; \
.set __pi_##x, x; \
.size __pi_##x, . - x; \
ENDPROC(x)
/*
* Emit a 64-bit absolute little endian symbol reference in a way that
* ensures that it will be resolved at build time, even when building a
* PIE binary. This requires cooperation from the linker script, which
* must emit the lo32/hi32 halves individually.
*/
.macro le64sym, sym
.long \sym\()_lo32
.long \sym\()_hi32
.endm
/*
* mov_q - move an immediate constant into a 64-bit register using
* between 2 and 4 movz/movk instructions (depending on the
* magnitude and sign of the operand)
*/
.macro mov_q, reg, val
.if (((\val) >> 31) == 0 || ((\val) >> 31) == 0x1ffffffff)
movz \reg, :abs_g1_s:\val
.else
.if (((\val) >> 47) == 0 || ((\val) >> 47) == 0x1ffff)
movz \reg, :abs_g2_s:\val
.else
movz \reg, :abs_g3:\val
movk \reg, :abs_g2_nc:\val
.endif
movk \reg, :abs_g1_nc:\val
.endif
movk \reg, :abs_g0_nc:\val
.endm
#endif /* __ASM_ASSEMBLER_H */